An Incremental Planning Strategy-Based Genetic Algorithm for Integrated Scheduling and Routing Optimization in Smart Manufacturing

Production scheduling has always been an important topic in smart manufacturing especially integrated with dispatching and routing of automatic guided vehicle or autonomous mobile robots, which is generally modeled as a nondeterministic polynomial-hard integer programming problem. However, most existing studies solve this combinatorial optimization problem through decomposition, namely dealing with scheduling and routing separately, or just assuming a fixed predefined distance matrix for routing, which sacrifices the solution quality for shorter time of decision making. As for metaheuristic solution for such integrated models, the need for effective and efficient searching rules is urgent, so that the heuristics can be more fit for specific purpose and scenarios. Thus, in this paper, an integrated scheduling and routing model is proposed oriented towards the optimization of production scheduling and fleet management in smart manufacturing environment. A genetic algorithm (GA) is used to solve this problem based on incremental planning (IP) strategy which makes effect through ensemble of two local search operators considering the time windows and alternative paths for every transportation task of AGVs. The effectiveness of the proposed method is validated through a simulation case study with horizontal comparison. Results showed that the IP strategy can improve the solution quality and traffic load imbalance of the factory layout, by reducing the makespan and working time of AGV by about 11.5% and 9% respectively. About 80% of the tasks obtained a shorter makespan by adopting an alternative path, showing the ability of traffic load balancing among layout elements when the problem scale increases.